← Global AI Policy Radar · Methodology
Jurisdiction Baseline · AI × Policy

AI Regulation in China.

At a glance
AI-Specific Law
No comprehensive law; layered AI rules apply
Core AI Rules
CAC rules cover generative, algorithms, deep synthesis
Provider Duties
File algorithms, assess security, label synthetic content
Active Enforcement
Filing campaigns, rectification orders, suspensions, and penalties

China has no comprehensive national AI law in force as of today. Instead, AI is regulated by targeted rules from the Cyberspace Administration of China covering generative AI, algorithms, deep synthesis, and synthetic-content labeling, alongside broader personal-information and network-data security laws. These rules apply mainly to public-facing internet services in China and may require content controls, user disclosures, algorithm filings, security reviews, data governance, and record retention, with regulators able to order corrections, suspend services, or impose penalties.

In force today6 instruments
Interim Measures for the Management of Generative Artificial Intelligence Services in forceAI-specific
The core national rule for public-facing generative AI services in China, including services offered through APIs. Providers must manage training-data legality and quality, protect personal information, prevent unlawful or harmful content, conduct security measures and complaint handling, and complete applicable algorithm filing and security-assessment procedures. Violations can lead to warnings, correction orders, or suspension, with penalties also imposed under the Cybersecurity Law, Data Security Law, and Personal Information Protection Law. ([cac.gov.cn](https://www.cac.gov.cn/2023-07/13/c_1690898327029107.htm?ref_source=www.aiclubs.cn&utm_source=openai))
Read the full text ↗ · 2023
LicensingDisclosureContent moderation
Provisions on the Administration of Internet Information Service Algorithm Recommendation in forceAI-specific
Applies to internet services using recommendation, ranking, search, filtering, scheduling, personalization, and certain generative or synthetic algorithms. Providers must establish algorithm-safety controls, disclose the use and basic principles of recommendation, offer a non-personalized or opt-out option, protect minors, control unlawful content, and file covered algorithms with the CAC. ([cac.gov.cn](https://www.cac.gov.cn/2022-01/04/c_1642894606364259.htm?utm_source=openai))
Read the full text ↗ · 2021
DisclosureContent moderationMinors
Unlock the full baseline

Get the remaining 4 instruments and the founder FAQ. Free: subscribe once and every briefing on this device unlocks.

Already subscribed? Unlock →

Provisions on the Administration of Internet Information Service Deep Synthesis in forceAI-specific
Governs deepfake and other deep-synthesis services, including synthetic text, images, audio, video, and virtual scenes. Providers must review users and content, add required identifiers, retain logs, manage security incidents, and complete applicable filing or security-assessment procedures. ([cac.gov.cn](https://www.cac.gov.cn/2022-12/11/c_1672221949354811.htm?eqid=b8c802050000bc5500000006649272b5&utm_source=openai))
Read the full text ↗ · 2022
DisclosureContent moderationSafety & security
Measures for the Labeling of Artificial Intelligence-Generated and Synthetic Content in forceAI-specific
Since September 1, 2025, covered providers must apply visible labels and embedded metadata to AI-generated or synthetic text, images, audio, video, and virtual scenes. Platforms must detect or rely on user declarations of synthetic content, add notices where appropriate, preserve specified logs for at least six months in some no-visible-label cases, and disclose labeling rules in user agreements. ([cac.gov.cn](https://www.cac.gov.cn/2025-03/14/c_1743654684782215.htm?utm_source=openai))
Read the full text ↗ · 2025
DisclosureContent moderationSafety & security
Personal Information Protection Law in force
This is the main privacy regime for AI training, fine-tuning, inference, profiling, and employee or customer data use. It requires a lawful basis, purpose limitation, data minimization, transparency, safeguards for sensitive personal information, rules for automated decision-making, individual rights, and controls on cross-border transfers. ([miit.gov.cn](https://www.miit.gov.cn/jgsj/zfs/fl/art/2022/art_515a4b20c12f430eab54bb4f56d89f56.html?utm_source=openai))
Read the full text ↗ · 2021
Data processingDisclosureSafety & security
Regulation on Network Data Security Management in force
Effective January 1, 2025, this administrative regulation operationalizes network-data security obligations for AI companies handling large, sensitive, or potentially important datasets. It requires data-security systems, classification and protection measures, incident response, important-data identification and reporting where applicable, and enhanced obligations for large platforms and important-data processors. ([cac.gov.cn](https://www.cac.gov.cn/2024-09/30/c_1729384452307680.htm?utm_source=openai))
Read the full text ↗ · 2024
Data processingSafety & securityDisclosure
Also worth noting
  • A comprehensive Artificial Intelligence Law remains under study and is included in China's 2026 legislative work plan, but no such law was in force as of August 6, 2026. ([big5.www.gov.cn](https://big5.www.gov.cn/gate/big5/www.gov.cn/zhengce/content/202605/content_7068345.htm?utm_source=openai))
  • The CAC's revised Internet Information Services Management Measures were still a consultation draft after the July 3, 2026 second public consultation, with no effective date stated. The draft would further consolidate platform, content, algorithm, and AI-generated-content duties. ([cac.gov.cn](https://www.cac.gov.cn/2026-07/03/c_1784822399677167.htm?utm_source=openai))
  • Monitor implementing guidance and technical standards on AI-generated-content labeling, algorithm filing, security assessment, and important-data classification because these often determine operational compliance even when the core rule is already in force.
Pending and recent legislation8 measures
CN 令第21号 Order No. 21, Interim Measures for the Management of Artificial Intelligence Anthropomorphic Interactive Services
Passed · 2026-07-15
This in-force Chinese administrative measure establishes rules for the development and provision of AI services that simulate human-like interaction. It aims to standardize applications, protect national security and the public interest, and safeguard the rights of individuals, o
CN 人社部发〔2026〕40号 Implementation Opinions on Accelerating the Development of “Artificial Intelligence + Human Resources and Social Security” Applications (Min
Advancing · 2026-07-08
This administrative policy document sets out measures to accelerate the application of artificial intelligence in human-resources and social-security services. It is jointly issued by the Ministry of Human Resources and Social Security, the National Development and Reform Commiss
CN 商建发2026年第89号 Implementation Opinions of the Ministry of Commerce and Seven Other Departments on Accelerating the Development of “Artificial Intelligence
Advancing · 2026-06-18
The eight departments set out measures to promote the integration of artificial intelligence into consumption, including retail, services, and consumer-facing business models. The document provides national policy guidance for expanding AI-enabled consumption and improving relate
CN 工信部通信〔2026〕121号 Ministry of Industry and Information Technology Notice on Issuing the Implementation Opinions on Innovative Development of “Artificial Intel
Advancing · 2026-06-10
The notice issues a 2026–2028 implementation plan to promote the integration of artificial intelligence with the information and communications sector. It sets out development and policy priorities for advancing AI-enabled innovation across the industry.
CN 国务院2026年度立法工作计划 State Council 2026 Legislative Work Plan (Comprehensive Artificial Intelligence Legislation Arrangements)
Introduced · 2026-05-11
The State Council’s 2026 legislative work plan includes accelerating comprehensive legislation to promote the healthy development of artificial intelligence. It identifies future AI legislation as a national legislative priority rather than enacting specific operative requirement
CN c_1779979789523320 Implementation Opinions on the Standardized Application and Innovative Development of Intelligent Agents
Advancing · 2026-05-08
A Chinese administrative policy document establishing guidance for the standardized application, governance, and innovative development of intelligent-agent systems. It is intended to support responsible deployment while promoting development of agent technologies.
CN 国能发科技〔2026〕34号 Notice Issuing the Action Plan for Promoting Two-Way Empowerment Between Artificial Intelligence and Energy (Guo Neng Fa Ke Ji〔2026〕No. 34)
Advancing · 2026-05-08
The National Development and Reform Commission, National Energy Administration, Ministry of Industry and Information Technology, and National Data Administration issue an action plan to promote the reciprocal integration of artificial intelligence and the energy sector. It addres
CN 教科信〔2026〕1号 Notice of the Ministry of Education and Four Other Ministries on Issuing the Action Plan for “AI + Education” (教科信〔2026〕1号)
Advancing · 2026-04-10
The five ministries jointly issue an action plan to advance the integration and application of artificial intelligence across education. It establishes national implementation priorities and calls for coordinated action by relevant departments and educational institutions.

Live from the Global AI Policy Radar.

Founder FAQAI law, regulators, what applies, what is next, penalties
Is there an AI-specific law in force here?

No. China does not yet have a comprehensive national AI law in force. AI is governed by a layered set of AI-specific rules covering generative AI, algorithms, deep synthesis, and synthetic-content labeling, supported by privacy and network-data security laws.

Who are the key regulators for AI?

The Cyberspace Administration of China, or CAC, is the central regulator identified in the current AI rules. It oversees areas including algorithm filings, generative AI, deep synthesis, synthetic-content labeling, and network-data security, while enforcement can also rely on the underlying cybersecurity, data-security, and personal-information laws.

Which rules apply to AI systems today?

Public-facing AI services, including API services, may be subject to the generative AI measures, algorithm recommendation provisions, deep-synthesis provisions, and synthetic-content labeling measures. The Personal Information Protection Law applies to training, fine-tuning, inference, profiling, and use of employee or customer data, while the Regulation on Network Data Security Management applies to relevant data-security activities. Requirements can include content controls, disclosures, algorithm filings, security assessments, data governance, incident response, and record retention.

What is coming next, and when?

A comprehensive national AI law remains a legislative work item rather than an enacted statute. The source gives no expected enactment date. The synthetic-content labeling measures have applied since September 1, 2025, and the network-data security regulation has been effective since January 1, 2025.

What are the enforcement and penalty risks?

Enforcement is active and administrative. Regulators publish filing lists, conduct platform campaigns, and can order rectification, suspend services, or issue warnings. Violations may also trigger penalties under the Cybersecurity Law, Data Security Law, and Personal Information Protection Law.

Not legal advice. For educational purposes only. AI-researched against official sources (2026-08-06), links verified.

Need a company-specific assessment? Explore Nomos →